mistocr 0.2.11__py3-none-any.whl → 0.2.12__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (10) hide show
  1. mistocr/__init__.py +1 -1
  2. mistocr/_modidx.py +1 -0
  3. mistocr/refine.py +8 -3
  4. {mistocr-0.2.11.dist-info → mistocr-0.2.12.dist-info}/METADATA +1 -1
  5. mistocr-0.2.12.dist-info/RECORD +11 -0
  6. mistocr-0.2.11.dist-info/RECORD +0 -11
  7. {mistocr-0.2.11.dist-info → mistocr-0.2.12.dist-info}/WHEEL +0 -0
  8. {mistocr-0.2.11.dist-info → mistocr-0.2.12.dist-info}/entry_points.txt +0 -0
  9. {mistocr-0.2.11.dist-info → mistocr-0.2.12.dist-info}/licenses/LICENSE +0 -0
  10. {mistocr-0.2.11.dist-info → mistocr-0.2.12.dist-info}/top_level.txt +0 -0
mistocr/__init__.py CHANGED
@@ -1 +1 @@
1
- __version__ = "0.2.11"
1
+ __version__ = "0.2.12"
mistocr/_modidx.py CHANGED
@@ -22,6 +22,7 @@ d = { 'settings': { 'branch': 'main',
22
22
  'mistocr.core.wait_for_job': ('core.html#wait_for_job', 'mistocr/core.py')},
23
23
  'mistocr.pipeline': {'mistocr.pipeline.pdf_to_md': ('pipeline.html#pdf_to_md', 'mistocr/pipeline.py')},
24
24
  'mistocr.refine': { 'mistocr.refine.HeadingCorrection': ('refine.html#headingcorrection', 'mistocr/refine.py'),
25
+ 'mistocr.refine.HeadingCorrections': ('refine.html#headingcorrections', 'mistocr/refine.py'),
25
26
  'mistocr.refine.ImgDescription': ('refine.html#imgdescription', 'mistocr/refine.py'),
26
27
  'mistocr.refine.add_descs_to_pg': ('refine.html#add_descs_to_pg', 'mistocr/refine.py'),
27
28
  'mistocr.refine.add_descs_to_pgs': ('refine.html#add_descs_to_pgs', 'mistocr/refine.py'),
mistocr/refine.py CHANGED
@@ -4,9 +4,9 @@
4
4
 
5
5
  # %% auto 0
6
6
  __all__ = ['prompt_fix_hdgs', 'describe_img_prompt', 'get_hdgs', 'add_pg_hdgs', 'read_pgs_pg', 'fmt_hdgs_idx',
7
- 'HeadingCorrection', 'fix_hdg_hierarchy', 'mk_fixes_lut', 'apply_hdg_fixes', 'fix_hdgs', 'ImgDescription',
8
- 'describe_img', 'limit', 'parse_r', 'describe_imgs', 'save_img_descs', 'add_descs_to_pg', 'add_descs_to_pgs',
9
- 'add_img_descs']
7
+ 'HeadingCorrection', 'HeadingCorrections', 'fix_hdg_hierarchy', 'mk_fixes_lut', 'apply_hdg_fixes',
8
+ 'fix_hdgs', 'ImgDescription', 'describe_img', 'limit', 'parse_r', 'describe_imgs', 'save_img_descs',
9
+ 'add_descs_to_pg', 'add_descs_to_pgs', 'add_img_descs']
10
10
 
11
11
  # %% ../nbs/01_refine.ipynb 3
12
12
  from fastcore.all import *
@@ -64,6 +64,11 @@ class HeadingCorrection(BaseModel):
64
64
  index: int
65
65
  corrected: str
66
66
 
67
+ # %% ../nbs/01_refine.ipynb 19
68
+ class HeadingCorrections(BaseModel):
69
+ "Collection of heading corrections returned by the LLM"
70
+ corrections: list[HeadingCorrection]
71
+
67
72
  # %% ../nbs/01_refine.ipynb 21
68
73
  prompt_fix_hdgs = """Fix markdown heading hierarchy errors while preserving the document's intended structure.
69
74
 
{mistocr-0.2.11.dist-info → mistocr-0.2.12.dist-info}/METADATA RENAMED
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: mistocr
3
- Version: 0.2.11
3
+ Version: 0.2.12
4
4
  Summary: Batch OCR for PDFs with heading restoration and visual content integration
5
5
  Home-page: https://github.com/franckalbinet/mistocr
6
6
  Author: Solveit
mistocr-0.2.12.dist-info/RECORD ADDED
@@ -0,0 +1,11 @@
1
+ mistocr/__init__.py,sha256=X4KG3FscE5AhbGbcdDDgdDC550CVpxNMwdNLcx6EQ7M,23
2
+ mistocr/_modidx.py,sha256=55k_EmMPrRmLa04qCE-_8lukJPCax66q5MALfpa8T4A,4154
3
+ mistocr/core.py,sha256=ohh2ru05gUKbIQCRHPMz_hw4ui39FtpoV3_S3n4bl_c,7592
4
+ mistocr/pipeline.py,sha256=k5rtr9ccxTgwQAzw533iMIL9qwiCofugWv0fhotLzaI,1639
5
+ mistocr/refine.py,sha256=hQg4ZYRTkPFEiCNNBU2ykgXFfVLCNAp2IuwOwHBVQ2k,12983
6
+ mistocr-0.2.12.dist-info/licenses/LICENSE,sha256=xV8xoN4VOL0uw9X8RSs2IMuD_Ss_a9yAbtGNeBWZwnw,11337
7
+ mistocr-0.2.12.dist-info/METADATA,sha256=MB_1D5vpqGUIA_jl-jNHj7E5LHJol6YGU8lw-ZCFj_4,8417
8
+ mistocr-0.2.12.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
9
+ mistocr-0.2.12.dist-info/entry_points.txt,sha256=HjAiHozobM-alm_6bTF-ehRr2DD3KYE9PgRngelONOY,36
10
+ mistocr-0.2.12.dist-info/top_level.txt,sha256=LelTYnSpSXLy1Hb1m2YP3gt8luwP-I8KV0NjP_ucdSs,8
11
+ mistocr-0.2.12.dist-info/RECORD,,
mistocr-0.2.11.dist-info/RECORD DELETED
@@ -1,11 +0,0 @@
1
- mistocr/__init__.py,sha256=_MLx4ac1juJPWEEiC9kMQISX3x3jFBr507jM2P_hxMg,23
2
- mistocr/_modidx.py,sha256=WTS9JpZdbrp2LghjhOV-CK0JYChHE4PzttgKfh7pTy4,4028
3
- mistocr/core.py,sha256=ohh2ru05gUKbIQCRHPMz_hw4ui39FtpoV3_S3n4bl_c,7592
4
- mistocr/pipeline.py,sha256=k5rtr9ccxTgwQAzw533iMIL9qwiCofugWv0fhotLzaI,1639
5
- mistocr/refine.py,sha256=H_IAF02k6CwBQdDJm9txknzUcTlz245zXitaHELX-P4,12791
6
- mistocr-0.2.11.dist-info/licenses/LICENSE,sha256=xV8xoN4VOL0uw9X8RSs2IMuD_Ss_a9yAbtGNeBWZwnw,11337
7
- mistocr-0.2.11.dist-info/METADATA,sha256=CZ9TO24sY0OT9B19KGPIuy3MF9T7B1G9qHaVBoEMRIk,8417
8
- mistocr-0.2.11.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
9
- mistocr-0.2.11.dist-info/entry_points.txt,sha256=HjAiHozobM-alm_6bTF-ehRr2DD3KYE9PgRngelONOY,36
10
- mistocr-0.2.11.dist-info/top_level.txt,sha256=LelTYnSpSXLy1Hb1m2YP3gt8luwP-I8KV0NjP_ucdSs,8
11
- mistocr-0.2.11.dist-info/RECORD,,